README.md

metadata

license: cc-by-nc-4.0
language:
  - en
  - zh
tags:
  - embodied-ai
  - robotics
  - egocentric
  - video
  - 1000-hours
pretty_name: CJET EGO DATASET
extra_gated_prompt: >-
  Please provide the following information to request access to this dataset. We
  will review your request within 2-3 business days.
extra_gated_fields:
  Applicant Entity Type:
    type: select
    options:
      - Company
      - Academic/Research Institution
      - Individual
      - Other
  Intended Use:
    type: text
  Contact Information:
    type: text
  Phone Number:
    type: text

CJET EGO DATASET (1000-Hour Embodied AI Dataset)

📖 Overview

The CJET EGO DATASET is a large-scale, egocentric video dataset for Embodied AI, open-sourced by Ningbo ChunJian Electronic Technology Co., Ltd. (ChunJian Tech). It contains approximately 1,000 hours of first-person perspective video data. This dataset aims to accelerate research and development in embodied intelligence, robotic vision, and behavior understanding.

🎬 Sample Preview

This preview is a single merged video: the left panel shows the left-eye RGB view, and the right panel shows the corresponding left-eye depth visualization.

_{Left: left-eye RGB | Right: left-eye depth (colorized for visualization)}

Region	Modality	Description
Left	Left-eye RGB	First-person left-eye color video
Right	Left-eye Depth	Corresponding depth map, colorized for preview

⚠️ License & Restrictions

This dataset is licensed under the CC BY-NC 4.0 (Attribution-NonCommercial 4.0 International) license.

• ✅ Permitted: Personal learning, academic research, and non-commercial model training.
• ❌ Strictly Prohibited: Any form of commercial use, reselling, or unauthorized integration into commercial products.
• Any research outputs using this dataset must explicitly credit Ningbo ChunJian Electronic Technology Co., Ltd. (ChunJian Tech).

📝 How to Request Access

This is a Gated Dataset. Due to the massive size of the data and the non-commercial license restrictions, you must submit an application and receive approval before downloading.

1. Ensure you are logged into your Hugging Face account.
2. Agree to share your contact information and fill out the Access Request Form on this page.
3. The form requires the following information:
   • Applicant Entity Type: Company, Academic/Research Institution, Individual, or Other.
   • Intended Use: Please describe your use case in detail (e.g., pre-training embodied foundation models, visual navigation research).
   • Contact Information: Email address, phone number, or other preferred contact methods.
   • Phone Number: Your contact phone number, including country code.
4. Once the administrator approves your request, you will receive an email notification, and you can then download the dataset via git or the Hugging Face datasets library.

📬 Contact Us

For any questions regarding the dataset or access requests, please contact:

• Company: Ningbo ChunJian Electronic Technology Co., Ltd.
• Hugging Face Organization: https://huggingface.co/ChunJianElectron
• Email: marketing@cjetec.com

🤝 Commercial Cooperation

Ningbo ChunJian Electronic Technology Co., Ltd. provides customized data collection services and welcomes other forms of commercial cooperation.

If you are interested in commercial partnerships or need tailored datasets, please contact us directly via email at marketing@cjetec.com.

🛠️ Usage

Once your access is approved, you can load the dataset as follows:

from datasets import load_dataset

# You must configure your HF token after your access request is approved
dataset = load_dataset(
    "ChunJianElectron/CJET_EGO_DATASET",
    token="your_hf_token"
)